As is well known in the art, a biometrics technique is generally referred to as a method of identifying persons based on their unique physical or morphological features. As various kinds of services through a network are being populated with the development of communication technique, studies on biometrics are actively made for a user authentication.
Such biometrics technique is an individual authentication technique using bio information, such as individual body features or behaviors, which will never be lost or have no need to be memorized. Moreover, since the individual bio information has to be directly input to a biometrics system for authentication, the biometrics technique has a merit of high security over existing technologies using passwords from the view of security. Accordingly, the biometrics technique comes into a spotlight as the next generation technique to replace the existing password and ID card-based individual authentication technologies.
However, the biometrics technique has a problem in that forged bio information, which may be input to a system, deteriorates reliability and security of bio-recognition. Particularly there is a difficulty in distinguishing real bio information from forged bio information by using a typical bio-recognition algorithm.
In addition, in case of conventional face recognition method, although a forged face may be easily distinguished from a non-forged face by using a thermal distribution of a face photographed by a thermal infrared camera, such a thermal infrared camera has a high cost to be applied into a biometrics system. Further, users can feel uncomfortable by a conventional face recognition method making a user to speak or move to input the users' bio information. Furthermore, although there has been proposed a method for determining whether or not a forged face is present by using a degree of variation of an eye area in a face image, this method has also a problem in that it has to premise movement or change of eyes in the face area.